ELITE and MASS DYNAMICS: the EAST CENTRAL EUROPEAN EXAMPLE Zsolt Enyedi and Fernando Casal Bértoa

ELITE AND MASS DYNAMICS: THE EAST CENTRAL EUROPEAN EXAMPLE Zsolt Enyedi and Fernando Casal Bértoa

Introduction In a paper written in 1995 entitled “What is different about post-communist party systems?” Peter Mair applied the method that he called „ex adverso extrapolation‟. He matched his knowledge of the process of consolidation of party systems in the West with what could be known at that time about Eastern European history, society and emerging party politics. The article predicted long term instability for the region. Most of the reasons for this instability, like the weakness of civil society, the so-called-triple transition or the fluid social structures were regarded as quasi self-evident at that time, and were rarely disputed afterwards. The article also employed a more specific argument, emphasizing the historical sequence of political modernization. Thinking in terms of the thresholds of political development one must acknowledge that Eastern Europe is a highly modernized region: it passed the stages of incorporation, mobilization, activation, politicization and contestation. The problem, as emphasized by Mair, was that contestation appeared as the last stage of the development, unlike in Western Europe, where it preceded democratization. In the new social and technological environment the mass party strategy is no longer available. The collective identities are fragile, he noted, not only because of the fluid social structures but also because organizational networks that have been so crucial for anchoring voters in the West, are substituted or at least dominated by top-down political parties in the East. The result is general alienation of voters from party politics. The article also claimed that parties in the region have no clear boundaries, they overlap with social movements and interest groups. One fundamental task of the present paper is to evaluate the assessments and predictions of Peter Mair. The second is to differentiate the picture of Eastern Central European politics by identifying differences across countries and across time. Next to classical characteristics of party systems this paper focuses on the patterns of party relations and on party system closure.

The big picture Electoral volatility

1 In many countries, as predicted by Mair, loyalty among the elites and masses to specific parties proved to be fragile. Electoral volatility continues to be well above the (rising) West European average, and the tendency is not declining. In order to calculate volatility, we use here Pedersen´s already classic index, which measures „the net change within the electoral party system resulting from individual vote transfers‟. In terms of coding rules, we adopt Bartolini and Mair´s rules, that is 1. Volatility figures are always rounded up from the second decimal point. 2. When two or more parties merge to form a new party, or when one or more parties merge with an existing party, the relevant electoral volatility is computed by subtracting the combined vote of the merging parties in the election immediately preceding the merger. 3. When a party splits into two or more parties, the relevant electoral volatility is computed by subtracting the combined vote of the new parties from that of the original party in the election immediately preceding the split […] 5. When the name of a party changes, volatility is computed as if it were the same party […] (1990: Appendix) Following Sikk (2005), and taking into consideration the organizational instability of political parties in Eastern Europe, we have considered as splits all those separations which derive from a decision of a minority (or a relevant member of the leadership) within the structure of a given party. Figure 1 displays the average volatility scores for the 11 East Central European countries studied. Although these results should be treated with caution due to the significant fluidity of voters in certain pairs of elections,1 two important points of immediate interest can be noted. First, and most obvious, East Central European electorates have stabilized in different ways and at different rates. Secondly, and notwithstanding all attempts at discrediting Lipset and Rokkan´s (1967) well-known „freezing‟ hypotheses (Shamir 1984; Crewe 1985), West European voters are far more loyal than their East European counterparts. On average, the difference is more than the double (i.e. 14 points). Moreover, and in clear contrast to what can be observed among the post-communist electorates, none of the long- established democracies surpasses the 15% threshold, with two obvious exceptions: namely, Italy and The Netherlands. In East Central Europe only Croatia has maintained an average volatility level below what Pedersen considered an “earthquake” election.

1 In Bulgaria, for example, volatility increased from 25 to 48 between 1997 and 2001, while in Poland it decreased from 25 to 8 between 2007 and 2011. In a similar vein, in Hungary electoral fluidity decreased up to 10 points before 2006, just to reach 33% (an increase of 25 points) in 2010 (see also Enyedi and Casal Bértoa 2011: 133-134).

2 Figure 1. Average electoral volatility in Eastern Europe (1990-2012) 40 35 30 25 20 15 10 5 0

Electoral volatility

Source: Enyedi and Casal Bértoa (2011), Casal Bértoa (2013) and own calculations.

The variance between elections rather than among countries shows an even more striking divergence: while out of 97 pairs of elections held in Western Europe, more than four/fifths (i.e. 82.5%) did not cross the 15% barrier, in post-communist Europe only 5 (out of 51) achieved more stable scores: namely, 7.7 in Poland (2011), 8.4 in Hungary (2006) and, to a lesser extent, 13 in Estonia (2011), 14.9 as well as 13.1 in Croatia (2007 and 2011, respectively). The contrast is even clear when we take into consideration that, while up to 40 of the Eastern European elections surpassed the 20% limit, only 3 of the 11 West European (other than the Dutch) earthquake elections did it so: namely, Italy in 1994 (36.7), Portugal in 1995 (a bare 20.7) and, more recently, Ireland in 2011 (26.7). Although the temporal trends are too complex to summarise here, three general findings are considered to be important. First, out of the eleven countries in which volatility has declined over time nine are in the West. Secondly, while all Eastern European countries (with the exception of Poland but including both Hungary and Latvia, in which electoral fluidity seemed to be declining) have undergone important electoral earthquakes at the most recent legislative elections, only Austria, Finland, Ireland and Spain have experienced a similar fate in the West. Last but not least, independently of the decade examined, West European electorates continue to be less fluid than their post-communist counterparts.2 Such findings should come as no surprise (Bielasiak 2005; Lane and Ersson 2007; Bågenholm and Heinö, 2013). Party stability

2 On average, the East-West divergence from one decade to the other has even increased: namely, i.e. from 16 points in the 1990s to 17 in the 2000s.

3 At the elections significantly more new parties appear and many of them have the realistic chance of gaining governmental power. Figure 2. Average number of new parties in Eastern Europe (1990-2012) 6

New parties

Source: Own calculations.

A comparison of the degree of party continuity across Europe reveals another instance of the clear divergence between the two parts (i.e. old and new systems) of the continent. More concretely, as follows from Figure 3, which ranks all 11 East Central European nations according to the average age of their significant parties (Dix 1992; Jin 1995; Tavits 2005),3 East European political parties have had to struggle in order to stay alive. This is clearly visible when we acknowledge that 19 of the 39 East European political parties were created after the beginning of the current century (against only 5 in the West – Italy excluded). More importantly, out of those 19 cases of recent party formation, up to 10 were created a couple of months ahead of the most recent legislative elections. In this context, the Latvian case is particularly dramatic, with 3 new parties (2 of them mergers of previous political forces) coming to the fore just before the elections. But in terms of stability of party labels there is considerable variance in the region. In a number of countries the party systems are dominated by parties that are more than 15 years old. Figure 3. Average age of parties in Eastern Europe (1990-2012)

3 Following the literature on the subject, I consider „significant‟ parties to be those that received at least 10% of the vote during the last parliamentary elections (Mainwaring and Scully 1995).

4 70

Party age

Source: Bugajski (2002) and official party websites.

In clear contrast, Western voters have been facing the same political options for the last two decades.4 This is more straightforward in countries such as Spain, Portugal, Cyprus, Malta, Luxembourg, Ireland, Sweden, and, until very recently, also Greece and the UK. There party politics have been led by the same political forces without exception. The only two exceptions to this general rule are Italy and, to a lesser extent, France. In the former case, and as already explained, the „explosion‟ of the old party system in 1994 brought about new political parties which continue to struggle organizationally. In a similar vein, French political parties have experienced an important organizational restructuring since 2002 (mainly on the right of the political spectrum) although, it should be noted, the main four ideological options (i.e. Communism, Socialism, Liberalism and Gaullism) have by and large continued to structure electoral voting. All in all, and if we exclude the extreme Italian case, up to 42 out of the 50 West European political parties included in this analysis were already active at the moment the Wall came down in November 1989: this means a survival rate of 84%. Even if we were to look only at the patterns of partisan formation and development, comparing East European parties after the year 2000 with their Western counterparts after 1990, the number of organizational splits and totally new parties in the East (14 – 7 each) clearly outnumbers the number of mergers (6), with 5 of them taking place in just two countries (i.e. Latvia and Romania), suggesting without doubt a de-structuring pattern, certainly absent in the West (where mergers – 5, equals both splits – 3, and party foundations

4 For obvious reasons, most East Central European countries could not count on previous democratic experiences. The Czech Socialists and Polish Peasants´ Party are, perhaps, the only exceptions.

5 from scratch – 2).5 All in all, the truth is that if, as Sartori put it many years ago „a structured party system can be defined […] as a […] system in which the major parties become “solid” and more “real” than the personalities‟ ([1968] 1990: 75, 77), it is obvious that in post- communist Europe this has not yet happened, and will not happen for some time (Mair 1997). Number of parties (electoral and parliamentary arenas) The number of parties stays high, but it declined somewhat (the correlation between year and fragmentation is -.14, sig. .05) and since meanwhile the fragmentation of West European party systems increased, on this aspect there is a slow convergence going on between the two halves of Europe (figure 5). In the West the fragmentation of the socio- political structures (especially of the working class and religious blocs), while in the East the impact of the restrictive electoral rules and the learning of politicians and voters seems to contribute primarily to this convergence. Figure 4. Changes of the effective number of parliamentary parties in Eastern Central Europe

4,6

4,4

4,2

3,8

3,6

3,4

3,2

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Source: Own calculations based on Gallagher (2012).

Figure 5. Average “effective” number of electoral parties in Eastern Europe (1990-2012)

5 This number could even be reduced to 1 if we were to consider the True Finns, appeared in 1995, as the direct heir of the Finish Rural Party, founded 36 years before.

6 8 7 6 5 4 3 2 1 0

ENEP

Source: Gallagher (2012) and own calculations.

Figure 6. Average “effective” number of legislative parties in Eastern Europe (1990-2012) 6

ENPP

Source: Gallagher (2012) and own calculations.

A first look at Figures 5 and 6, which displays the average „effective‟ number of electoral/legislative parties in the region, reveals that, with the exception of Hungary and Bulgaria, highly fragmented party systems constitute the norm in the Eastern part of the European continent. Indeed, while on average the latter adhere to extreme pluralism (ENPP =5.6; ENPP=4.1), most of their Western European counterparts show limited pluralism (ENEP=4.6; ENPP=3.8). This is not to say, however, that in Western Europe fractionalized party systems cannot be found: for example, Finland, The Netherlands or Italy, where an

7 important reduction in the number of parties could be observed after the last elections due to the confluence of the main political groupings in two big parties (namely, the People of Freedom and the Democratic Party). In Belgium the „real‟ number of parties, all currently in government, is 6. Interestingly enough, and despite the fact that East Central European party systems are more fragmented than their Western European counterparts, the differences – as mentioned above – are not so striking: just 0.3 points. As explained elsewhere (Casal Bértoa, 2013), this is mainly a function of the most disproportional electoral systems adopted in the post- communist region (LSq. = 7.1 in EU 10+1 vs. 5.2 in EU17). Party membership Concerning party organizations a number of scholars, primarily the Mair-disciple Ingrid van Biezen, have confirmed the weakness of organizational encapsulation and the elitist nature of parties, but with the important difference vis-à-vis Western Europe that the party in public office is weaker than the party in central office. Figure 7. Party membership in Eastern Europe (2010/2011) 7,0

6,0

5,0

4,0

3,0

2,0

1,0

0,0

M/E

* It excludes micro-states (i.e. Malta and Luxembourg) Note: Biezen et al. (2012)

According to Biezen et al. (2012), and notwithstanding an important decrease in the last decades, the total party membership as percentage of the electorate (i.e. M/E) continues to be on average higher in the West than in the East (figure 7). In particular, while in Western Europe only 5 countries do not display a M/E ratio higher than 4 percent, in the East only Bulgaria, Estonia and Slovenia reach such figure.

8 At the same time Margit Tavits has shown that better organization provides a large advantage in the electoral competition even in this region. Indeed, the party that Peter Mair described in the article as the best example of the conscious rejection of grass-roots involvement, the Hungarian Fidesz, has developed one of the most formidable organizational structures in Europe. Later, in a separate work with Petr Kopecky, another disciple, Mair also showed that the state in Eastern Europe is colonized by the parties, and that parties have some unusual possibilities to exploit their power position due to the many non-routine activities of the state like privatization. In some relevant dimensions his references to „lack of systemness‟ may be considered to be exaggerated. Most importantly, as far as the ideological positions of electorates and parties are concerned the Eastern European map, although different from the Western European one, is not necessarily more chaotic. Many scholars, including Lisbet Hooghe, identified relatively simple and meaningful ideological landscapes in the region. Figure 8. Average ideological polarization in Eastern Europe (1990-2012) 0,6

0,5

0,4

0,3

0,2

0,1

Dalton´s Index

Source: Döring and Manow (2012)

Using Dalton´s (2008:8) polarization index (PI), which measures „the distribution of parties along the [commonly known] Left-Right scale‟,6 Figure 8 displays the degree of ideological polarization characterizing each of the East Central European democracies here studied. On the one hand, while ideological polarization is high in both the Czech Republic

6 PI is calculated according to the following formula: PI = {[Σ(vi)*([xi-x]/5)²]½, where vi is the proportion of votes of the ith party, xi refers to its left-right score, and x represents the average party system score on the left- right scale (Dalton 2008:9). The index goes from 0 (non-polarized) to 1 (polarized).

9 and Hungary, Slovenia, Bulgaria, Slovakia and Romania are among the least polarized East Central European party systems, with the other four post-communist countries somehow in the middle. On the other, it follows from the figure above that region matters very little. A comparison of the average level of polarization in both parts of the continent – the difference is just 0.03 – clearly confirms a pattern of convergence regarding this issue. All in all, the truth is that both the continuous decline and progressive disappearance of traditional communist and/or fascist political parties in Europe have helped to erode the Sartorian distinction between „centripetal‟ and „centrifugal‟ dynamics.7 This is not to deny, however, a more general tendency towards bi-polar competition between „opposing‟ party blocs observed in both Eastern and Western Europe (see below, but also Enyedi and Casal Bértoa 2011; or Mair 2006). Still, and with some exceptions, the former has not prevented the formation of parliamentary coalitions between previously totally inimical camps. In the post-communist region, Poland (2007) and Romania (2008) are cases in point. The one point where Mair‟s assessment is not valid concerns the boundaries of parties and their hierarchical position vis-a-vis social movements. While on this aspect we have no easily accessible numerical indicators, the literature (e.g. Webb and White 2007) indicates that parties have fairly clear identities and the movements and interest groups tend to be relegated to the margins of the political systems. Voter turnout The existence of strong partisan oppositions and the dominant position of parties does not entail, however, high level of popular involvement. The level of electoral turnout is low and is declining.

Figure 9. Average voter turnout in Eastern Europe (1990-2012)

7 In this point, it is clear that the evaporation of the Communist bloc, which led to the „triumph‟ of the neo-liberal paradigm, together with the erosion of the traditional socio-political cleavages in a more secular and globalized world, have definitely helped to reduce the level of ideological polarization in most Western European polities.

10 90 80 70 60 50 40 30 20 10 0

Voter turnout

Note: Voting is compulsory in Belgium, Cyprus, Greece and Italy (until 1994).

Bloc politics Mair predicted in the paper a particularly conflictual politics in the region. According to his logic open markets, large stakes, high uncertainty need to lead to adversarial behavior. While the polarization figures do not necessarily prove that the region would sharply differ from the West, he was substantially right. In the East European new democracies the rules of the game are often changed by the winners, and the losers are marginalized in a more aggressive way than in the West. In this regard, East and West show diverging patterns. This observation should be perhaps linked to the findings by Mair and Kopecky concerning patronage: The penetration of parties into the state may be deep, but rent-seeking tends to be less systematically organized than in the West. Eastern Europe doesn‟t look like a hospitable ground for cartel party development, at least as far as the cooperation among established parties is concerned. With a few expectations, the lack of trust among the main players and their fears of populist challengers constrain the degree of cooperation. It is interesting to note that Peter Mair typically interpreted cross-party cooperation in a negative way, as far as Western Europe, while he regarded the “culture of compromise and accommodation” as a necessary condition for Eastern European consolidation. One may criticize this as lack of consistency, but probably Peter Mair simply sensed very well that in different stages of democracy different degrees of cooperation are needed. The analyzed article suggested long term instability. And yet, the author still expected that passing of time will make a difference. Mair emphasized that one needs time for

11 coalitional norms to develop and for collective identities to crystallize. From today‟s vantage point one can say that he was too cautious. He wrote, for example, “Much like Germany and Italy in the postwar decades, and Portugal and Spain in the last twenty years, the post- communist democracies will also eventually assimilate to the pattern set by the established democracies.” As the figures cited above indicate, the expectations towards a gradual process of maturation and stabilization are not (yet) met.

Party system closure: conceptualization and operationalization Among the factors that contribute to the institutionalization of party systems only a few are under the direct influence of the parties themselves. Electoral volatility, fragmentation, and polarization, for example, depend on a host of social and political phenomena, including the socio-structural environment, the media, the institutional environment, ideological traditions, mass organizations, economic climate, etc. The aspect of party systems that reflects more accurately party agency and the cooperative and competitive relations among parties, is the aspect of coalition patterns. Mair (1997:206, 2001, 2007) drew attention to three related phenomena: the form of government alternation, the degree of the access to office and the innovative nature of the coalition formulae. Together they show whether a party system is closed, predictable or open, inchoate. The first dimension is alternation in government. As far as this aspect is concerned three patterns exist: wholesale-, partial-, and non-alternation. In the first case the incumbent government leaves the office in its entirety, and is replaced by a completely different group of parties. The second option materializes when the new cabinet is a mixture of parties from inside and outside of the previous government. The third possibility is marked by a complete absence of alternation, as the same party or parties remain in exclusive control of government over an extended period of time. The second major component of the model, governing formulae, shows whether the partisan composition of the governments is innovative or familiar. Familiarity prevails if there are stable groups of parties that tend to govern together. If there is a tendency towards previously unseen party compositions of cabinets, the system is considered to be open, innovative. The final component is access to government. The relevant difference on this aspect is between, on one hand, systems in which all parties have the opportunity to participate in government and, on the other, polities in which some parties are permanently excluded.

12 Closed governments consist exclusively of parties that governed in the past, the open ones include, or are dominated, by novices. By assessing these three factors one can determine whether a party system is closed or open. Party systems are considered to be open when (1) the alternations of governments tend to be partial, (2) no stable configuration of governing alternatives exists and (3) the access to government is granted to all relevant parties, including the newly established ones. It is closed when (1) alternations of governments are wholesale or none, (2) the governing alternatives are stable over a long period of time, and (3) governments are confined to a narrow circle of parties. The concept of part system closure has already inspired a number of party scholars. Some of them used the original measures proposed by Mair (e.g. Toole, 2000; O´Dywer, 2006;), while others developed new, dichotomous or continuous indicators, and Mair himself kept changing the measures. The original conceptualization suggested a dichotomous operationalization (e.g. Henjak, 2003; Stoychev, 2008): governmental formulae are either innovative or familiar, the governments either include newcomers or not, alternation is either wholesale (i.e. total or none) or partial. Müller-Rommel (2005) applied a score of 0 to no- alternation, innovation and access, 1 to wholesale alternation, familiar formula and closed access and 2 to partial alternation. Enyedi and Casal Bértoa (2011) assigned 0 to partial alternation, innovation and access, and 1 to wholesale alternation, familiar formula, and closed access. Some studies (Mair, 2007; Casal Bértoa and Mair, 2012) used continuous variables to grasp the degree of change. For capturing alternation in government Mair (2007) applied Pedersen´s (1979) index of electoral volatility to the ministerial turnover (calling it the index of governmental alternation, IGA). Thereby IGA was computed by adding the net change in percentage of ministers (including the prime minister) gained and lost by each party in the cabinet from one government to the next, and then dividing the result by two. Concerning the familiarity/innovation of formulae Mair‟s continuous indicator relied on the number of innovative governments as a percentage of all governments in a particular period. Finally, for access Casal Bértoa and Mair created an index of openness (IO) which “measures the weight new parties have in a particular cabinet as well as the weight such governments (with new parties) enjoy in the party system as a whole” (2012:101). IO is calculated by dividing the number of new governing parties by the total number of governing parties. The operationalizations summarized above have merits, but also considerable weaknesses. Dichotomous measures are obviously crude indicators. Larger the new parties

13 joining the government, larger the rupture with the past. Therefore, in principle, continuous variables should be preferred. The exact construction of such variables requires, however, additional theoretical and empirical justifications. As demonstrated below, the existing continuous variables are all flawed in one way or another. Our effort is aimed at correcting their flaws and suggesting new indices that follow a uniform logic. This common logic implies taking the percentage of minister-changes as the basis for all the three variables8. One of the fundamental issues with Mair‟s original operationalization is the way how government-change is defined. Traditionally, as summarized for example by Müller and Strøm (2000:12), three criteria are used: (1) a change in the partisan composition of the government coalition; (2) a change of the head of government; and (3) new elections. Contrary to this standard, for calculating the indices of closure Mair defined individual governments exclusively in terms of partisan composition. In this understanding a government survives until its partisan composition is kept intact. If the same coalition continues after an election Mair has one entry in his dataset for the pre- and post-election period, not two. We agree with Mair‟s decision not to consider changes in prime ministers as relevant end-points. The significance of the replacement of the PM typically does not reach the weight of a new election or the political relevance of a coalition break-up. PM-changes which are not results of elections and are not accompanied by party-composition-changes tend to concentrate in those countries where position of the premier is less important. Elections, on the other hand, are defining political events. At each new election the citizens and the political class face a dilemma: to continue with the old patterns or to innovate. Elections always give the possibility for the reconfiguration of the party landscape. Continuity in these cases is therefore noteworthy. The challenge arising from a PM‟s resignation is typically managed by a small group of politicians within the dominant party, while a national parliamentary election “tests” the entire society. Therefore we consider elections as signaling change of government. Additionally, one can record “no-alternation” only if one considers elections as defining criteria of the units of measurement. As reviewed above, according to Mair‟s original conceptualization a system should be considered to be closed whenever a particular government stays in power continuously. This plausible logic is, however, not well reflected in the scores of closure as cases of no alternation are by definition non-existent in Mair‟s dataset.. To give an example, India was a particularly closed system during its first two decades of democracy according to the Mair‟s definition but using his operationalization this

8 Similarly to Mair we consider ministers and not ministries, and we count them as equally relevant.

14 cannot be demonstrated as India hasn‟t experienced a single change of partisan composition during this period. Those of the indicators in the literature which characterize times-spans do so by averaging the scores of consecutive governments. As a result no difference is made between short- and long-lived governments and particular time-points are not evaluated. We suggest to assign figures to years and to take the mean of the years included into the analyzed time period as the period-specific closure-figure, multiplying the cabinet characteristics by the number of years the government lasted. If the country had more governments in a specific year then their average is used to characterize the particular year. As far as other aspects of the counting rules are concerned, we largely follow Mair. “Grand coalitions”, i.e. coalitions bringing together the most relevant parties (e.g. Kiesinger‟s 1966 or Merkel‟s 2005 cabinets) are included into our dataset, but “exceptional”: acting, in exile, provisional, ad interim, care-taker, presidential, national union (or ecumenical), technocratic, non-partisan governments, are not. To name a few, the latter categories include the Poincaré (1926), Brünning (1930), Churchill (1940) or, more recently, the Monti (2010) or Pikrammenos (2012) governments. The characteristics of these cabinets can tell us little about the way in which political parties normally interact. Finally, note that when counting ministers within a particular government we consider to which parties, and not to which electoral coalitions, they belong. We start our counting from the moment a so-called founding government is established, provided that a country is both (1) democratic (i.e. a Polity IV score of 6 or higher) and (2) independent. The year when a country ceases to be democratic is excluded, except in those cases when the authoritarian transition was preceded by a change of government or by elections taking place during that year. As mentioned above, Mair (2007) suggested ministerial volatility as the indicator of alternation. Higher volatility was expected to reflect high closure. The implicit logic behind this operationalization is that the most closed party systems are characterized by a bipolar pattern of competition undergoing regular and complete alternation. Consider, however, a stable Social Democratic-Green coalition which is enlarged from time to time with a five percent strong Left-Libertarian party. In this case continuity prevails and change/surprise is marginal. But because the Pedersen index amounts to a mere five percent, using Mair‟s index one would mistakenly conclude that the system is particularly open, unpredictable. The bias in the operationalization stems from the fact that only wholesale alternation is recognized as closure, no-alternation is not. Correcting for this bias means that each party-

15 composition change needs to be placed on a continuum between total9 and no alternation. The degree of stability is indicated by the distance from that endpoint of the continuum which is further away. If 85% of a government is changed then the case is almost a case of total alternation. The figure 85 expresses well this high degree of stability. But if only 10% of the ministers belong to new parties then continuity is even higher. The numeric value assigned to the latter case should be 90. In this way one can assign to each government-change a figure that ranges between 50 and 100, the former indicating openness, the latter indicating closure. Concerning alternation and access (see below), the construction of percentage-based measures is relatively simple. With the familiarity of government formulae there is more complication. It is probably not accidental that no continuous indicator exists in the literature that could be applied to single governments. In extreme cases the construction of the variable is easy. If the very same combination of parties has already governed together in the analyzed period, there is a 100 percent familiarity. If the government is based on an entirely new combination of parties, the familiarity figure is 0. In the more complicated case when only part of the new government replicates a previous government we suggest to use the percentage of the ministers who belong to the „familiar‟ part of the previous government. The contrast needs to be made with the previous government that is most similar in terms of the list of parties to the analyzed government. A cabinet of Communists, Socialists , Greens and Populists must be compared to the government with Socialists , Greens and Populists , and not to the more dissimilar one containing Communists and Socialists . What shall we do with single party governments? If the party has never governed before, the adequate closure figure is 0, the formula is innovative. But what if the party was already part of a coalition before and now is governing alone? Here we propose that the percentage of the previous coalition partners should be subtracted from 100. If the Social Democrats, who gave 90 percent of the ministers, decide to continue without the Greens, who dominated only one tenth of the government, then closure of formula is high, 90. In the opposite case innovation prevails, since a small, marginal party turns into the sole government-party. Closure then equals 10. As a result, the index of government formula ranges between 0 and 100.

9 Note that total alternation can go together with openness if the new government is composed of entirely new actors. In these (rare) instances the alternation figure will be 100, the formula and access figures will be both 0, the composite closure figure will 33.

16 The existing continuous indicator of access to government (the ratio between the new governing parties and all governing parties) also needs to be altered, in two ways. First, the values need to be reversed so that they express the degree of closure, in line with the other two components. Second, in order to capture the magnitude of change, the percentage of ministers who are controlled by particular parties should be taken into account, again similarly to the other two indicators. The proposed measure is therefore the percentage of old governing parties. Old governing parties are considered to be all those parties which have already been constituent parts of a previous government. The indicator again varies between 0 and 100. A country‟s composite closure score will equal the average score of the three variables. Since the minimum value of alternation score was not 0 but 50, for calculating the average the value of alternation was linearly transformed: subtracting 50 and multiplying with 2.10 The closure index can be used to characterize a particular year or a longer period. In the first instance, if, for example, completely new parties form a government, closure is 0. But if one evaluates the entire stock of government formations during e.g. 20 years, then this 0 comprises only one twentieth of the overall score. When showing the trajectories of party systems we use the latter, the cumulative index, that is we incorporate all data prior of the analyzed time point. Party system closure: East vs. West First let us compare the region with the old member states in 2012. The closure index shows a 10 point difference between the two regions (figure 10). The Western Europe average is close 95%, the Eastern European one is at 85%. The Eastern figures are moving closer to the Western ones, but very slowly: in the period between 1990 and 1997 the average was 79,9, between 1998 and 2002 19,1, between 2003 and 2007 81,6 and between 2008 and 2012 83,1.

10 Researchers may consider using our index in two further alternative forms. The first is using the standardized versions (Z scores) of the original variables, the second is projecting them onto a 0 to 1 scale. Sometimes it is beneficial to center the variable around the mean, as the Z scores do, but then one must consider that the scores will depend on the specific sample used. The second alternative is particularly suitable for regression analyses. Here we opted for the 0 to 100 scale because it can be interpreted as percentage of stability and it clearly communicates how much stability prevails over change.

Figure 10. Party system closure in Eastern Europe (1990-2012)

100

LIT HUN SLO SLK CZE CRO EST POL LAT ROM BULG EU 17 EU11

Source: Own calculations

Figure 11. Government alternation in Eastern Europe (1990-2012)

100

LIT HUN CRO CZE SLK POL SLO EST LAT BULG ROM EU 17 EU11

Source: Own calculations

Figure 12. Access to government in Eastern Europe (1990-2012)

100

LIT EST HUN CRO CZE SLK SLO POL LAT ROM BULG EU 17 EU11

Source: Own calculations

Figure 13. Governing formula familiarity in Eastern Europe (1990-2012)

LIT SLO HUN CZE SLK EST CRO LAT POL ROM BULG EU 17 EU11

Source: Own calculations

The difference between the two regions appears on all the three components of closure (figure 11, 12 and 13). But the shift towards closure is due to the cumulative stability of

19 governmental formulae and to the limited access to government, the patterns of alternations have not changed significantly (figure 14). Figure 14. Changes in party system closure in East Central Europe (1990-2012)

alternation 75 formula access

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 1994

Intra-regional variance Trends in the dimensions of party systems The data presented so far allow the identification of the intraregional differences. Taking the two decades of multiparty democracy into account Hungary, Croatia, Slovenia, the Czech Republic and Romania belong to the more institutionalized group, while Latvia, Lithuania, Poland to the more inchoate one, with Bulgaria, Estonia, and Slovakia oscillating, depending on the time period or the indicator examined. Figure 15. Changes in the number of effective parliamentary parties (1990-2012)

10 BULG CRO CZE 8 EST HUN 6 LAT LIT POL 4 ROM SLK 2 SLO

1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012

Source: Own calculations

20 As it follows from the figure above and the table below, fragmentation increased in Bulgaria, Lithuania, while it decreased in Estonia, Hungary, Latvia, Poland, Croatia and Slovenia (and, possibly, Romania). In Slovakia and Czech Republic it fluctuates. Table 1. Association between year and fragmentation (Pearson coefficients and levels of significance) ENPP BUL CRO CZ EST HUN LAT LIT POL RO SK SLO

,641** -,958** -,388 -,527* -,826** -,573* ,937** -,474* -,448 ,008 -,850** ,002 ,000 ,091 ,017 ,000 ,010 ,000 ,030 ,082 ,975 ,000 21 12 20 20 22 19 19 21 16 20 19 Source: Own calculations

There is variance among the countries in the region also in terms of party system closure. First we have checked the correlation between time (years) and the year-specific closure figures (table 2). According to the data closure increased in Latvia, Poland, Slovakia (and non-significantly in Estonia and Hungary)11, and it has decreased in Czech Republic, Lithuania and Romania. Alternation became more closed in Bulgaria, Hungary, Poland, Latvia and more open in Czech Republic, Estonia and Lithuania. As far as formula is concerned there is movement towards closure in Latvia, Poland and Slovakia, and shift towards openness in the Czech Republic, Romania and Estonia. Access became more restricted in Latvia Poland and Slovakia, the only country where there is a tendency towards openness is the Czech Republic. The Czech Republic stands out as a case that opened up on all fronts, while Latvia and Poland are the ones that became more structured in all three aspects. Table 2. Association between year and party system closure and its components: alternation, formula and access (Pearson coefficients and levels of significance) Closure BUL CRO CZE EST HUN LAT LIT POL RO SK SLO

,303 ,124 -,799** -,391 ,380 ,809** -,533* ,678** -,554* ,704** ,177 ,182 ,700 ,000 ,088 ,081 ,000 ,019 ,001 ,026 ,001 ,469 21 12 20 20 22 19 19 21 16 20 19 alternation BUL CRO CZE EST HUN LAT LIT POL RO SK SLO

,635** ,451 -,909** -,562** ,552** ,479* -,853** ,500* ,010 ,322 ,270 ,002 ,141 ,000 ,010 ,008 ,038 ,000 ,021 ,970 ,167 ,264 21 12 20 20 22 19 19 21 16 20 19 Formula BUL CRO CZE EST HUN LAT LIT POL RO SK SLO

,105 -,213 -,746** -,501* ,026 ,802** -,386 ,651** -,854** ,750** ,197 ,652 ,506 ,000 ,024 ,908 ,000 ,102 ,001 ,000 ,000 ,418 21 12 20 20 22 19 19 21 16 20 19 Access BUL CRO CZE EST HUN LAT LIT POL RO SK SLO

11 When calculated so that only election years are taken into account the association becomes significant.

,163 ,181 -,517* ,196 ,304 ,802** -,131 ,751** ,123 ,672** ,177 ,481 ,573 ,020 ,407 ,168 ,000 ,593 ,000 ,651 ,001 ,469 21 12 20 20 22 19 19 21 16 20 19 Source: Own calculations

To grasp a different aspect of closure lets examine the graphs indicating how the cumulative indices of closure changed across the years (figures 15 and 16). The first graph presents the entire period, while the second leaves out the first five years from the trajectory of the party systems. We should focus on the second graph because changes during the very first years will have a disproportionately large impact on the cumulative scores - as the system had no possibility yet to accumulate inertia. Figure 16. Changes in the cumulative closure scores (1991-2012)

100

90 BULG

80 CRO CZE 70 EST HUN 60 LAT LIT 50 POL ROM 40 SLK SLO 30

1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011

Source: Own calculations

As can be observed above, the general tendency is one towards closure, but there were larger shifts towards openness in Bulgaria in 2001 and 2009, in Croatia in 2008, in Estonia in 2002-2003, in Latvia in 2002, in Lithuania in 2000-2001, in Poland in 2005-2006, and in Slovakia in 2002. The Czech party system seems to gradually deinstitutionalize. The consecutive shocks of Tosovsky‟s centrist cabinet (1998), Zeman‟s leftist cabinet (2002), then Topolanek‟s government including the Greens (2007) and Nekas‟s cabinet including Top09 and VV (2010) weakened the system.

22 Figure 17. Changes in the cumulative closure scores without the first five years of multiparty democracy (1997- 2012)

100

95 BULG CRO 90 CZE EST 85 HUN LAT 80 LIT POL 75 ROM SLK SLO 70

1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Source: Own calculations

Latvia, Lithuania and Poland have the least closed systems throughout the period. Latvia shows a gradual process towards closure, but it started from a very low base. During the last years the governmental arena of Lithuania also increased its predictability, but it has still not reached the level where it was during the mid 1990s. Poland showed signs of settling down during the turn of the millennium, but the 2005-2006 upheaval pushed the country back to the least institutionalized group. Estonia and Croatia form medium-closed systems, with shift towards higher closure during the last years. The Bulgarian trajectory is more perplexing, with periods of high stability, and then sudden collapses. Romania and the Czech Republic used to have the most predictable governmental arena but during the last years this stability gradually eroded, although they still belong to the more institutionalized group. Finally, Slovenia and Hungary exhibit a level of closure that is similar to the West European averages, even if both experienced recently some changes in the patterns of government formation.

Patterns of party systems While the continuous variables discussed above provide for a relatively simple way for characterizing the trajectories, not all aspects of party systems can have a numerical representation. The profiles of the party systems should include the ideological character, the number and the relationships among the principal electoral alternatives, the poles of the party

23 systems. Below we present a very short description of the 11 party systems in terms of the changing polarities and alliance structures. The first case, the Slovenian party system started off fragmented, with a rapidly emerging pivotal party. The left- and right-wing parties played a secondary role in government coalitions dominated by the liberal centre. The relational and ideological structuration of the party system remained low, but this loose form of structure proved to be enduring, lasting until the collapse of the Liberal Democrats in 2004. The second phase is characterised by a more bipolar logic (primarily pitting conservatives against social- democrats, SDS vs. SD). The adversarial atmosphere, together with the failed crisis- management, led in 2011 to an early election and to the victory of a new center-left party at the polls. The government formed after the 2011 election lasted only one year. The tripolar pattern had a much shorter life-span in Hungary, and barely lasted until 1994. Even during this period, when the socialists (MSZP), conservatives (led by MDF) and the liberals (led by SZDSZ) offered distinct ideological alternatives, the discourse and the political preferences tended to be structured in a bipolar way. Between 1994 and 2010 a two- bloc-, and almost two-party-pattern prevailed: the leftist MSZP, and the right-wing Fidesz jointly received close to 90 per cent of the vote. The particularly aggressive, uncompromising attitude of the blocs towards each other engendered a centrifugal pattern of competition. In 2010 the two blocs were complemented by a third, half-bloc, the extreme right, but the fundamental logic of the competition remained bipolar. The extreme right does not (yet) play a role in the competition for government. In the Czech Republic after a short transition period during which the communists were challenged by a gradually fragmenting anti-communist bloc, a pattern emerged that is similar to the latest Hungarian phase: two-and-a-half blocs in terms of relations and ideology but bipolar logic of competition for government, dominated by CSSD and ODS (but note that these two parties are typically supported by half as many voters as Hungary‟s leading parties). As opposed to Hungary the extremist half-bloc is leftist. The small parties in the middle play an important role in government building but not in terms of the definition of the alternatives. In the Estonian party system the competition originally unfolded between moderate and technocratic reformers and nationalist (but also pro-West) anti-communists. The second pattern, the one that is still in place, is a multi-polar one, with four almost equal sized actors: conservatives (led by Isamaa), liberal technocrats (led by Reform), the leftish, moderately populist, pro-Russian Centre Party (Keskerakond) and the social-democrats (SDE). The coalition-possibilities are open, but most of them include the centrist liberals.

24 The trajectory of the Romanian party system is divided into four phases. The first configuration was bipolar, pitting the post-communist socialists against a cluster of parties that was dominated by conservatives and liberals. Between the two blocs operated the small but significant social-liberal Democrats and on the right flank of the system the nationalists, who cooperated during the nineties with the socialists. In the second formula the two major alternatives remained the same, but the nationalists evaporated and the Democrats, combining populist and centre-right elements, gained significance. The third pattern reflected the further ascendance of Democrats who became a major pole of party competition. The three parties (PDS, PNL and PDL, together with a few minor satellites) formed an almost perfect triangle, leaving open the possibility of cooperation on all fronts. In the last stage socialists and liberals formed a united front against the Democrats, re-establishing the bipolar pattern. The first Slovak party system was anchored by the presence of a sizable nationalist force that was opposed by the colorful bloc of liberals, conservatives and social-democrats. The second system that emerged around the millennium continued the bipolar pattern, transforming the nationalist camp into a more standard leftist pole (Smer). Latvia, as demonstrated above, is characterized by high degree of fragmentation both in terms of individual parties and clusters of parties. It has a stable leftist-Russian pole, led by Harmony, and a complex right wing camp, as of 2013 led by the populist Unity. This is the country in our sample that comes closest to a „no-system‟. To the extent that the opposition between the Russian and the Latvian parties is taken as a defining aspect, one can speak of a bipolar system. Lithuania started with a two-bloc competition, dominated by the attitude towards the communist heritage, pitting LDDP against Homeland Union. This simple logic was disturbed at the end of the century by the emergence of, first, the liberals, then of populists, but these forces ended up aligning with the left, and therefore the bipolar logic re-emerged. In its first phase the Polish system was tripolar, but the position of liberals, situated between socialists and conservatives, weakened rapidly. The second phase forms a two-and-a-half pattern, the main rivalry unfolding between nationalist conservatives (PiS) and conservative liberals (PO), with the socialists forming a half-pole. The Croatian party system was dominated to such an extent by HDZ, a conservative- nationalist party, during the early 1990‟s , that the country approached a one-party state. Since the government had both left-wing and right-wing critics one could consider Croatia as a center-based dominant party system. In the second part of the 1990‟s a very solid two-pole pattern emerged with the SDP challenging HDZ from the left.

25 Finally, the Bulgarian party system, as many other countries in our sample, was also originally structured by the anti-communist cleavage. The two-bloc structure (socialists, led by BSP, and conservatives, led by SDS) lasted almost a decade, producing a centrifugal and bipolar structure. This transparent logic was complicated by the emergence of a centrist, liberal force (NDSV). After the collapse of this party the bipolar logic remerged, but the traditional right wing parties were replaced by a new, conservative-populist bloc under the leadership of GERB. Table 3. The consecutive configurations of party systems in East Central Europe Bulgaria bipolar bipolar Croatia center-based bipolar dominant party system Czech Republic bipolar two-and-a-half Estonia bipolar multi (four)-polar Hungary tripolar bipolar two-and-a-half Latvia no system/bipolar? Lithuania bipolar bipolar Poland tripolar two-and-a-half Romania bipolar two-and-a-half tripolar Slovakia bipolar bipolar Slovenia center-based bipolar multipolar

The qualitative assessment of the patterns of party systems (table 3) largely corresponds with the quantitative analysis. Systems that have a transparent conflict structure tend to have more closed governmental arena. At the same time it is also clear that bipolarity does not guarantee stability and predictability.

Conclusions One obvious conclusion from this review of various aspects of party systems is that party politics in the region has not consolidated yet. Fragmentation declined moderately, the closure of the governmental arena slightly increased, and there are fewer new parties in the parliament. But electoral volatility hardly changed and turnout decreased. The closure of the governmental competition increased due to minor increase in the familiarity of coalition formulae and the modest decline in access to government office, but the patterns of alternation have not changed. Hungary, Croatia, Slovenia, the Czech Republic and Romania belong to the more institutionalized group, while Latvia, Lithuania, Poland to the more inchoate one, while

26 Bulgaria, Estonia, and Slovakia oscillate, depending on the time period or the indicator examined. Fragmentation is high in Latvia, Lithuania, Slovenia, Czech Republic, Estonia and Romania, but in the other coutries is around the Weste European average, and in Croatia and Hungary it is well below it. Fragmentation increased in Bulgaria, Lithuania, while it decreased in Estonia, Hungary, Latvia, Poland, Croatia and Slovenia. In terms of the closure of the governmental arena, one can find the most open systems of the continent in this region: Latvia, Lithuania and Poland. But Hungary and Slovenia are not less predictable than the West European systems. Closure increased in Latvia, Poland, Slovakia, and it has decreased in Czech Republic, Lithuania and Romania. Alternation became more closed, that is more bipolar, in Bulgaria, Hungary, Poland, Latvia and more open, that is more partial, in Czech Republic, Estonia and Lithuania. As far as formula is concerned there is movement towards familiarity in Latvia, Poland and Slovakia, and there is a shift towards innovation in the Czech Republic, Romania and Estonia. Access became more restricted in Latvia, Poland and Slovakia. In terms of closure Bulgaria appears as a special case: between elections little changes and the government parties do not cooperate with opposition parties after the next election, but the country is characterized by high unpredictability in terms of governmental formulae and, especially, access. Romania and the Czech Republic used to have the most predictable governmental arenas and they still belong to the more closed systems but the tendency towards decline is unmistakable. Systems that have a transparent conflict structure tend to have more closed governmental arena, but bipolarity does not guarantee stability and predictability.

References (to be completed) Bartolini, S. and Mair, P. (1990): Identity, Competition, and Electoral Availability: the Stabilization of European Electorates 1885–1985. Cambridge: Cambridge University Press Biezen van, I., Poguntke, T. and Mair, P. (2012): “Going, Going…Gone? The Decline of Party Membership in Contemporary Europe”, European Journal of Political Research, v. 51, n. 1, pp. 24-56 Casal Bértoa, F. (2013): “Post-communist Politics: on the Convergence or Divergence of East and West”, Government and Opposition, v. 48, n. 3 Casal Bértoa, F. and Enyedi, Z. (2010): “Party System Closure: Conceptualization, Operationalization, and Validation”, DISC Working Paper Series, n. 11

27 Casal Bértoa, F. and Mair, P. (2012): “Party System Institutionalization across Time in Post- Communist Europe” (with Peter Mair), in Ferdinand Müller-Rommel and Hans Keman (eds.) Party Government in the New Europe. Routledge Dalton, R.J. (2008): “The Quantity and the Quality of Party Systems”, Comparative Political Studies, v. 20, n. 10, pp. 1-22 Dix, R. (1992): “Democratization and the Institutionalization of Latin American Political Parties”, Comparative Political Studies, v. 24, n. 1, pp. 488-496 Döring, H. and Manow, P. (2011) Parliament and government composition database (ParlGov): An infrastructure for empirical information on parties, elections and governments in modern democracies. Available at: http://dev.parlgov.org/data/. Last accessed on June, 20th 2012 Enyedi, Z. and Casal Bértoa (2011): “Patterns of Inter-Party Competition (1990-2009)”, in Paul G. Lewis and Radosław Markowski (eds.) Europeanizing Party Politics? Comparative Perspectives on Central and Eastern Europe. Manchester: Manchester University Press Gallagher, M. (2012): “Electoral Systems” Website. Available at: http://www.tcd.ie/Political_Science/staff/michael_gallagher/ElSystems/Docts/ElectionIndices. pdf. Last accessed on June, 20th Gallagher M., Laver, M., and Mair, P. (2011): Representative Government in Modern Europe. Boston: McGraw-Hill Laakso, M., and Taagepera, R. (1979): “„Effective‟ Number of Parties. A Measure with Application to West Europe”, Comparative Political Studies, v. 12, n. 1, pp. 3-27 Lewis, P.G. (2006): “Party Systems in Post-communist Central Europe: Patterns of Stability and Consolidation”, Democratization, v. 13, n. 4, pp. 562-583 Lipset, S.M. and Rokkan, S. (1967): “Cleavage Structures, Party Systems and Voter Alignments: an Introduction”, in Seymour M. Lipset and Stein Rokkan (eds.) Party Systems and Voter Alignments. New York: Free Press Mainwaring, S., and Scully, T. (1995): Building Democratic Institutions: Party Systems in Latin America. Stanford, CA: Stanford University Press Mair, P. (1997): Party System Change. Approaches and Interpretations. Oxford: Clarendon Press Mair, P. (2006): “Party System Change”, in Richard S. Katz and William Crotty (eds.), Handbook of political parties. London: Sage Morlino, L. (1998): Democracy between Consolidation and Crisis: Parties, Groups, and Citizens in Southern Europe. Oxford: Oxford University Press

28 Müller, W.C. and Fallend, F. (2004): “Changing Patterns of Party Competition in Austria: From Multipolar to Bipolar System”, West European Politics, v. 27, n. 5, pp. 801-835 Pedersen, M.N. (1979): “The Dynamics of European Party Systems: Changing Patterns of Electoral Volatility”, European Journal of Political Research, v. 7, pp. 1-26 Sartori, G. (1976): Parties and Party Systems. A Framework for Analysis, Volume I, Cambridge: Cambridge University Press Tavits, M. (2005): “The Development of Stable Party Support: Electoral Dynamics in Post- Communist Europe”, American Journal of Political Science, v. 49, n. 2, pp. 283- 298 Toole, J. (2000): “Government Formation and Party System Stabilization in East Central Europe”, Party Politics, v. 6, n. 4, pp. 441-461